wood-fired-tasks 1.12.0

package/AGENTS.md

@@ -0,0 +1,112 @@
+# AGENTS.md
+
+Owner: Repository maintainers
+
+First-read navigation hub for any AI coding agent working on `wood-fired-tasks`. Vendor-neutral. Points at deeper docs; does not duplicate them. The full contract is in [docs/AGENT_CONTEXT.md](docs/AGENT_CONTEXT.md).
+
+## What is wood-fired-tasks
+
+`wood-fired-tasks` is a single-repo TypeScript task-tracking system that exposes the same underlying data through three surfaces and one notification channel:
+
+- A **Fastify REST API** (`src/api/`) for HTTP clients.
+- An **MCP server** (`src/mcp/`, `@modelcontextprotocol/sdk`) so AI agents can read and mutate tasks as tools.
+- A **Commander CLI** named `tasks` (`src/cli/`) for terminal users.
+- A **Slack bolt** notifier (`src/slack/`) that posts task lifecycle events.
+
+All three surfaces sit on a shared core: typed Zod schemas (`src/schemas/`), services (`src/services/`), repositories (`src/repositories/`), and a **better-sqlite3** database (`src/db/`) with umzug migrations. Tests are vitest; lint is biome; Node ≥ 22 ESM throughout. Run `npm run dev` for the REST API, `npm run mcp:dev` for the MCP server, `npm run cli -- <command>` for the CLI. Everything else (build, lint, tests, migrations) is `npm run …`. There is no separate frontend, microservice, or background worker — one process per surface, one shared SQLite file, one TypeScript codebase.
+
+## Read-next by intent
+
+Pick your intent, read the files in order. Files marked `(reserved)` are slots defined by the contract but not yet on disk in this milestone.
+
+| Intent | Read in this order |
+|---|---|
+| docs-only change | `docs/AGENT_CONTEXT.md` → `CONTRIBUTING.md` → the doc you are editing |
+| API change (REST routes) | `docs/API.md` → `src/api/routes/` → an existing test in `src/api/__tests__/` → `docs/INTERFACES.md` |
+| MCP tool change | `docs/MCP.md` → `src/mcp/tools/` → `src/mcp/__tests__/` → `docs/INTERFACES.md` |
+| CLI change | `docs/CLI.md` → `src/cli/commands/` → `src/cli/__tests__/` |
+| Schema / status / enum change | `src/schemas/` → matching `src/services/` or `src/repositories/` → API/MCP/CLI surface that exposes it |
+| Database migration | `src/db/migrations/` → `src/db/migrate.ts` → `src/db/__tests__/` → `docs/ARCHITECTURE.md` |
+| Slack change | `docs/SLACK.md` → `src/slack/` → `slack-app-manifest.yml` |
+| Test-only fix | failing test file → the unit under test → `vitest.config.ts` |
+| Release / docs update | `docs/RELEASE.md` → `CHANGELOG.md` → `package.json` |
+| Deploying your fork to production | `docs/SETUP.md` (Self-hosting and upgrades) → `deploy/install.sh` → `deploy/upgrade.sh` → `docs/RELEASE.md` (Migration expectations) |
+
+For per-surface change recipes (18 task shapes with files / tests / docs), see [`docs/NAVIGATION.md`](docs/NAVIGATION.md).
+
+## Essential commands
+
+| Purpose | Command |
+|---|---|
+| Install dependencies | `npm ci` |
+| Build (and typecheck) | `npm run build` |
+| Lint | `npm run lint` |
+| Focused test file | `npx vitest run path/to/file.test.ts` |
+| Focused test by name | `npx vitest run -t "name pattern"` |
+| Full test suite | `npm test` |
+| Local REST API server | `npm run dev` |
+| Local MCP server (stdio) | `npm run mcp:dev` |
+| Remote MCP server (HTTP) | `npm run build && npm run mcp:remote` |
+| Run a migration | `npm run migrate` |
+| Run the CLI in-tree | `npm run cli -- <args>` |
+| Full quality gate | `npm run quality` |
+
+`npm run build` is the project's typecheck — there is no separate `tsc --noEmit` script.
+
+## High-risk / handle-with-care areas
+
+Treat these as off-limits unless your task explicitly requires touching them.
+
+- `data/`, `*.db`, `*.db-wal`, `*.db-shm` — gitignored SQLite files; commonly hold **real task data**. Never commit, never delete blindly, never assume schema from byte layout — read `src/db/migrations/` instead.
+- `dist/` — gitignored build output; regenerated by `npm run build`. Never hand-edit.
+- `coverage/`, `reports/`, `.stryker-tmp/` — gitignored test artifacts. Safe to delete; never commit.
+- `.env`, `.env.local` — gitignored; carry credentials and tokens. Never paste contents into chat, logs, or commits. Use `.env.example` as the template.
+- `.gitleaks-report.json` — gitignored secret-scan output. Local-only.
+- `/bin/` — gitignored client install artifacts. `bin/tasks.cmd` may contain a **baked-in API key** from a local install; do not commit, share, or read its contents into context.
+- `.planning/`, `.claude/`, `.codex/`, `.agents/`, `.bug-smash-*.md` — gitignored workspace dirs. Not part of the shipped repo; do not rely on them.
+
+## Deeper docs
+
+| File | One-line purpose |
+|---|---|
+| [README.md](README.md) | Product-level overview, install, quickstart |
+| [CONTRIBUTING.md](CONTRIBUTING.md) | Human contributor workflow, commit and PR rules |
+| [docs/AGENT_CONTEXT.md](docs/AGENT_CONTEXT.md) | Authoritative contract for agent-facing files |
+| [docs/API.md](docs/API.md) | REST API reference |
+| [docs/MCP.md](docs/MCP.md) | MCP tool reference |
+| [docs/CLI.md](docs/CLI.md) | CLI reference |
+| [docs/SETUP.md](docs/SETUP.md) | Local setup, env, install |
+| [docs/SLACK.md](docs/SLACK.md) | Slack integration reference |
+| [docs/RELEASE.md](docs/RELEASE.md) | Release process |
+| [docs/CODE_QUALITY_ROADMAP.md](docs/CODE_QUALITY_ROADMAP.md) | Quality roadmap |
+| [docs/ONBOARDING_SMOKE.md](docs/ONBOARDING_SMOKE.md) | Onboarding smoke test — 7 probe scenarios for fresh agents |
+| [docs/REPO_MAP.md](docs/REPO_MAP.md) | Compact repo tree with per-directory ownership |
+| [docs/ARCHITECTURE.md](docs/ARCHITECTURE.md) | System one-pager, data flow across surfaces |
+| [docs/WORKFLOWS.md](docs/WORKFLOWS.md) | Canonical command recipes (build, test, lint, run) |
+| [docs/INTERFACES.md](docs/INTERFACES.md) | Inventory of REST routes, MCP tools, CLI commands (counts verified by CI) |
+| [docs/NAVIGATION.md](docs/NAVIGATION.md) | Task-oriented "if you want to do X, read these files" index |
+| [docs/ONBOARDING_SMOKE.md](docs/ONBOARDING_SMOKE.md) | Repeatable onboarding smoke test (scripted + manual) |
+| [docs/TROUBLESHOOTING.md](docs/TROUBLESHOOTING.md) | Operator recovery runbook: boot failures, wrong/stale DB, safe backup/restore |
+| [.agent-context.json](.agent-context.json) | Machine-readable manifest of the files above |
+
+## Task-orchestration skills (`/tasks:*`)
+
+This repo ships agent skills under `skills/tasks/` that automate the
+plan→execute→audit loop over a wood-fired-tasks project. They install to
+`~/.claude/commands/tasks/` via `install.sh`. The orchestration set:
+
+| Skill | Status | One-line purpose |
+|---|---|---|
+| `/tasks:decompose` | OPERATIONAL | Break a project-level goal into 8–25 leaf tasks (or a dependency DAG) ready for an executor. Planner only — never executes. See [`skills/tasks/decompose.md`](skills/tasks/decompose.md) and the design at [`docs/tasks-decompose-design.md`](docs/tasks-decompose-design.md). |
+| `/tasks:loop` | OPERATIONAL | Drain a FLAT-topology backlog sequentially. See [`skills/tasks/loop.md`](skills/tasks/loop.md). |
+| `/tasks:loop-dag` | OPERATIONAL | Drain a DAG-topology backlog wave-by-wave in parallel. See [`skills/tasks/loop-dag.md`](skills/tasks/loop-dag.md). |
+| `/tasks:audit` | OPERATIONAL | Retroactively grade a completed loop run. See [`skills/tasks/audit.md`](skills/tasks/audit.md). |
+
+Typical flow: `/tasks:decompose` a goal → run `/tasks:loop` (FLAT advisory)
+or `/tasks:loop-dag` (DAG advisory) → `/tasks:audit` the run. Decompose
+plans and hands off; the executors run; the auditor grades — three separate
+orchestrators by design.
+
+## Vendor neutrality
+
+`AGENTS.md` is the authoritative entry point for every agent, regardless of vendor. Vendor-specific files (`CLAUDE.md`, `.cursor/`, `.gemini/`, `.codex/`, any future `.<vendor>/`) MAY exist but MUST be either thin pointers back here or vendor-only configuration (slash commands, MCP client wiring, tool allow-lists). They MUST NOT carry unique project facts; if you find one that does, move the content into the authoritative tier (`AGENTS.md` or `docs/**`) and replace the vendor file with a pointer. The full boundary rules are in [docs/AGENT_CONTEXT.md §6](docs/AGENT_CONTEXT.md).

package/CHANGELOG.md

@@ -0,0 +1,271 @@
+# Changelog
+
+All notable changes to this project are documented in this file.
+
+The format is based on [Keep a Changelog 1.1.0](https://keepachangelog.com/en/1.1.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+Each release groups changes under `Added`, `Changed`, `Fixed`, and `Security`.
+Security-relevant items (auth, secret handling, network exposure, dependency
+vulnerabilities, supply-chain pinning) are always called out under `Security`.
+
+## [Unreleased]
+
+## [v1.12] - 2026-05-25
+
+First public open-source release. OSS-launch readiness and CI sharding work landed since v1.11.
+
+### Added
+- `docs/TROUBLESHOOTING.md` operator recovery runbook (boot failures, wrong/stale DB, safe backup/restore); linked from AGENTS.md, README, and the docs index (task 355).
+- Tool-count drift regression test for public docs (task 260).
+- `.env.example` aligned with documented config (task 259).
+
+### Changed
+- License relicensed from ISC to MIT for OSS launch (task 253).
+- SECURITY.md rewritten for the actual TypeScript/Fastify stack (task 254).
+- OSS package metadata + npm `files` allowlist (task 255).
+- Installer split into local vs. remote MCP paths; `--api-key` flag deprecated (still parsed and honored, emits a deprecation warning — prefer `WOOD_FIRED_TASKS_API_KEY`, the per-user secret file, or the interactive prompt) (task 258).
+- Stryker mutation tests sharded across 4 parallel CI jobs; threshold raised to 75; workflow timeout extended (tasks 250, 252).
+- Shipped systemd unit (`deploy/wood-fired-tasks.service`) now orders after `network-online.target` so OIDC discovery doesn't crash-loop the service on a cold boot; `StartLimitBurst` raised to 5 (task 353).
+- `docs/MCP.md` now recommends the remote (REST) variant as the single-writer default, warns that the local direct-SQLite variant silently serves stale data, and documents a launcher-wrapper that keeps the API key out of client config (task 356).
+
+### Fixed
+- Test runs cleanly clean up `task.status_changed` listeners between runs (task 257).
+
+## [v1.11] - 2026-05-21
+
+Coverage ratchet, CLI bare-spot tests, and workflow engine cascade-rollback fix.
+
+### Added
+- Vitest coverage thresholds ratcheted past 80/80/70 (task 249).
+- Test coverage for remote MCP rest-client + register-tools (task 249).
+- Test coverage for CLI bare-spot commands `completed`/`completions`/`db-check`/`doctor` (task 249).
+- Test coverage for CLI interactive prompts (task 249).
+- Test coverage for CLI formatters + JSON output (task 249).
+
+### Fixed
+- Workflow engine no longer leaks phantom events during cascade rollback (task 244).
+
+## [v1.10] - 2026-05-21
+
+Pagination envelope, schema-derived CLI types, and a new MCP completion-report tool.
+
+### Added
+- `completion_report` tool on the remote MCP server (task 245).
+
+### Changed
+- List endpoints documented with paginated envelope shape (task 248).
+- CLI `TaskResponse` derived from the server Zod schema instead of being hand-maintained (task 246).
+- CLI shell-completions command list derived from the Commander registry (task 247).
+
+### Fixed
+- Dropped duplicate `metadata.range` field in `tasks completed --json` output.
+
+## [v1.9] - 2026-05-20
+
+Documentation sweep for OSS launch — README, CONTRIBUTING, CHANGELOG, API/CLI/MCP docs, and de-personalized fixtures/scripts.
+
+### Added
+- `CHANGELOG.md` with backfill of recent security-relevant changes (task 217).
+- GitHub issue and PR templates (task 232).
+- CONTRIBUTING.md expanded for external contributors (task 229).
+
+### Changed
+- README badges, refreshed test counts, canonical URLs, branding sweep (task 238).
+- SETUP, Slack, data model, env vars, and architecture docs expanded (task 214).
+- MCP docs corrected: tool count, remote server, events resource (task 220).
+- CLI docs document 6 previously-missing commands; fixed `claim` JSON shape (task 223).
+- API docs corrected: endpoint surface, `/health` & swagger, env, filters (task 227).
+- Test fixtures de-personalized with generic names (task 237).
+- Deploy scripts de-personalized via `WFT_SERVICE_USER` (task 236).
+
+## [v1.8] - 2026-05-20
+
+Pre-OSS hardening — test coverage, mutation testing, and performance regression
+gates added as part of the open-source audit.
+
+### Added
+- Vitest `bench` suite for hot-path perf regression (task 212).
+- v8 coverage reporter + CI threshold gate (task 199).
+- Stryker mutation score enforcement in CI (task 203).
+- Cross-platform install-script smoke tests + linters (task 202).
+- OpenAPI snapshot for contract drift detection (task 207).
+
+### Changed
+- Expanded fast-check property tests over service invariants (tasks 200, 209).
+- CLI snapshot + real-binary end-to-end tests (tasks 208, 211).
+- Migrations verified via up→down→up schema snapshots (task 201).
+
+### Fixed
+- `install.sh` `TEMP_FILES` expansion under `set -u` on bash 3.2.
+- CI shellcheck scoped to `install.sh`; tests excluded from `tsc` build.
+- SSE fan-out and Slack request-signing regression coverage (tasks 205, 206).
+
+### Security
+- Generalized personal/internal examples in docs and tests pre-OSS.
+
+## [v1.7] - 2026-05-20
+
+Security audit remediation release — concentrated mitigation of pre-OSS
+security review findings.
+
+### Changed
+- Default HTTP bind moved to `127.0.0.1` (task 188).
+- `limit` / `offset` pagination on list endpoints (task 192).
+
+### Fixed
+- DELETE comment validates task ownership (task 191).
+- Slack `subscribe` event types validated against allowlist (task 198).
+
+### Security
+- Untracked `.planning/` and internal agent dirs (task 186 prep).
+- Documented `.env` handling, scrubbed working-tree secrets (task 187).
+- Admin trust model documented + per-key labels (task 189).
+- Removed live API key from events MCP resource markdown (task 196).
+- Removed hardcoded internal LAN IP (task 190).
+- SSE map stores hashed API-key fingerprint, not raw key (task 194).
+- `install.sh` escapes API keys via `jq -n --arg` (task 195).
+- All GitHub Actions pinned to commit SHAs (task 197).
+
+## [v1.6] - 2026-05-20
+
+First wave of security-audit fixes plus client-package / remote-MCP work.
+
+### Added
+- Remote MCP proxy server backed by the REST API for thin clients (quick-6).
+- Client package with `setup.bat`/`uninstall.*` scripts and bundled CLI.
+- `updated_after` / `updated_before` filters on `GET /tasks` (task 100).
+- Completion-report dashboard (task 97).
+
+### Changed
+- `list_tasks` MCP response trimmed to compact summaries by default.
+- Migration name normalization prevents `.ts` / `.js` mismatch crashes.
+
+### Fixed
+- Startup retries for transient SQLite errors in MCP entry point.
+- Migrations serialized with exclusive SQLite transaction lock (quick-5).
+- `SlackService` logger type widened to `FastifyBaseLogger`.
+- Client install handles spaces in paths and PowerShell 5.1.
+
+### Security
+- Reduced unauthenticated public surface, bounded SSE connections (task 185).
+- Hardened API-key auth against weak keys + brute force (task 182).
+- Upgraded vulnerable prod deps + CI `npm audit` gate (task 181).
+- Validated FTS search input, mapped syntax errors to 400 (task 183).
+- Secured API-key handling across all installer scripts (task 184).
+
+## [v1.5] - 2026-02-18
+
+Slack integration milestone.
+
+### Added
+- Slack Bolt integration (`SlackService`) with channel subscriptions and
+  `subscribe`/`unsubscribe` command surface (phases 23–26).
+- `tasks` slash-command router with task/project/dep/comment/subtask/health
+  subcommands and Block Kit formatters.
+- `UserIdentityCache` for Slack user ID → display name resolution.
+- `SlackNotifier` subscribed to EventBus with retry and per-channel error
+  isolation.
+
+### Changed
+- Slack token config validated both-or-neither via Zod.
+
+## [v1.4] - 2026-02-17
+
+Reliability, UX polish, and CI quality gates.
+
+### Added
+- `tasks doctor`, `tasks stats`, `tasks db-check` diagnostics (phase 19).
+- `tasks backup` CLI command + `backlogged` status (phase 18).
+- Property-based tests for `CycleDetector` and status transitions (phase 20).
+- Stryker mutation testing; knip unused-dep detection; GitHub Actions CI.
+- Request IDs, MCP `traceId` logging, reduced SSE buffer.
+
+### Changed
+- Phase 17 Core Reliability work landed across 4 plans.
+- Phase 21 UX polish: spinner, color utils, shell completions.
+
+### Fixed
+- Sweep no longer reverts `done`/`closed` tasks (quick-4).
+- CLI `apiRequest` handles HTTP 204; SSE returns 400 on missing `Accept`.
+
+### Security
+- systemd unit gains resource limits + security hardening (phase 22).
+
+## [v1.3] - 2026-02-14
+
+Workflow engine, atomic task-claim protocol, and SSE event stream.
+
+### Added
+- `EventBus` with generics + Task/Project CRUD emissions (phase 14).
+- `GET /api/v1/events` SSE endpoint, `SSEManager` with buffering.
+- Atomic claim protocol (CAS + `BEGIN IMMEDIATE`), `claim_task` MCP tool,
+  `tasks claim` CLI, stale-claim auto-release service (phase 15).
+- `WorkflowEngine` parent auto-complete with cascade depth + dependency
+  auto-unblock (phase 16).
+
+### Changed
+- `@fastify/sse` API usage corrected for v0.4.0.
+
+## [v1.2] - 2026-02-14
+
+Installers, Claude Code skills, and documentation.
+
+### Added
+- Bash and PowerShell installer scripts (phase 13).
+- Claude Code skill files: `create-task`, `log-bug`, `my-work`, `show-task`,
+  `pick-up`, `search`, `add-comment`, `done`, `blocked`, `project-status`
+  (phase 12).
+- `README.md`, `docs/SETUP.md`, `docs/API.md`, `docs/CLI.md`, `docs/MCP.md`.
+- E2E regression suite and skill-file validation tests.
+
+### Fixed
+- Redirected Umzug logger to stderr so MCP stdio output stays clean.
+
+## [v1.1] - 2026-02-13
+
+CLI expansion, MCP tool surface, JSON mode.
+
+### Added
+- Project, dependency, comment, subtask, and health CLI commands
+  (phases 08-01 through 08-05).
+- MCP project tools, `list_subtasks` task tool, and `check_health` health
+  tool (phase 09).
+- Global `--json`, `--no-input`, `--force` flags and `@clack/prompts`
+  interactive prompts (phase 07).
+- `NO_COLOR` environment variable support.
+
+### Fixed
+- HTTP client only sends `Content-Type` when the request has a body.
+- `dotenv` stdout contamination suppressed so JSON output stays clean.
+
+## [v1.0] - 2026-02-13
+
+Initial tagged release — REST API, CLI, MCP server, SQLite backing store,
+and the task/project/dependency/comment/subtask domain model.
+
+### Added
+- Fastify REST API: task + project CRUD, auth, error handler, health,
+  OpenAPI docs (phase 02).
+- CLI with API client + task `create`/`list`/`update`/`delete`/`show`
+  commands (phase 03).
+- MCP server with task CRUD tools + error conversion (phase 04).
+- systemd service unit + production entry point with graceful shutdown
+  (phase 05-01).
+- SQLite backup/restore scripts with cron schedule (phase 05-02).
+- Task hierarchy (subtasks), dependency service, comments, time estimates
+  (phase 06).
+
+[Unreleased]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.11...HEAD
+[v1.11]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.10...v1.11
+[v1.10]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.9...v1.10
+[v1.9]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.8...v1.9
+[v1.8]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.7...v1.8
+[v1.7]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.6...v1.7
+[v1.6]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.5...v1.6
+[v1.5]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.4...v1.5
+[v1.4]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.3...v1.4
+[v1.3]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.2...v1.3
+[v1.2]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.1...v1.2
+[v1.1]: https://github.com/Wood-Fired-Games/wood-fired-tasks/compare/v1.0...v1.1
+[v1.0]: https://github.com/Wood-Fired-Games/wood-fired-tasks/releases/tag/v1.0

package/CLAUDE.md

@@ -0,0 +1,21 @@
+# CLAUDE.md
+
+> See [AGENTS.md](AGENTS.md) for the canonical, vendor-neutral agent entry
+> point. This file exists so Claude Code automatically discovers the project
+> context — it is an **adapter** with no unique facts.
+
+## What to read first
+
+1. [AGENTS.md](AGENTS.md) — first-read navigation hub.
+2. [docs/AGENT_CONTEXT.md](docs/AGENT_CONTEXT.md) — the contract that
+   defines what counts as canonical agent context here.
+3. [.agent-context.json](.agent-context.json) — machine-readable manifest.
+
+## Maintenance
+
+- This file MUST stay a thin pointer. The contract is in
+  [docs/AGENT_CONTEXT.md](docs/AGENT_CONTEXT.md) §6.2.
+- The freshness check (`npm run agent-context:check`) verifies that every
+  link target exists; the manifest classifies this file as `authority: adapter`.
+- Project-specific Claude Code config (slash commands, hooks, MCP client
+  wiring) lives in `.claude/` (gitignored). Do not mirror facts here.

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Wood Fired Games
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.